At the University of Chicago, several residents and fellows felt that PPI infusions were being over-ordered in patients with upper gastrointestinal (GI) bleeding. Clinicians would often reflexively order 72-hour PPI infusions prior to endoscopy but fail to make necessary changes later, said Vineet M. Arora, MD, MAPP, FACP, associate professor of medicine and assistant dean for scholarship and discovery at the university's Pritzker School of Medicine.
“A lot of times, especially post-endoscopy, the GI team did notice that even though the lesions were not high-risk, patients were kept on their 72-hour IV infusion,” she said.
After confirming with pharmacy staff that about half of the infusion orders were indeed inappropriate, the trainees initiated a “Skip the Drips” intervention in July 2015.
How it works
The quality improvement project began with a system change, delivered through the electronic health record (EHR). Physicians were prompted to select an appropriate indication for their PPI infusion orders, and pharmacists provided oversight by tracking monthly usage of PPI infusions.
Training came in the form of education directed toward internal medicine and ED residents and faculty. Rather than providing a lecture, the team used “guerrilla tactics,” plastering work rooms and resident areas with brochures explaining the proper indications for ordering PPI infusions, said Dr. Arora. “The brochure was endorsed by the chief quality officer, as well as our GI fellowship director, so it helped that it had that institutional approval,” she said.
A cultural component of the intervention was delivered by designated faculty champions, who helped spread the word. Soon the “skip the drips” message was ubiquitous. “On one ICU round, one of the resident champions for this project was rotating in the ICU, and another resident was like, ‘Well, shouldn't we skip the drips?’” Dr. Arora said. “People were mentioning it without him having to say it.”
After the intervention (July 2015 to April 2016), an average of 15% of PPI infusion orders were inappropriate in patients admitted for upper GI bleeding, compared to an average of 50% in the preceding year (July 2014 to June 2015), according to results published as a research letter in November 2017 by JAMA Internal Medicine.
“I was surprised to see such a substantial drop, knowing that we didn't do so much education,” said Dr. Arora, the paper's senior author. “It really spoke to me about the need to embed those system changes and how the EHR could actually help enable you to do the right thing.” As inappropriate use of PPI infusions went down, appropriate use went up: Postendoscopy PPI orders followed established guidelines 95% of the time after the intervention, compared to 83% in the preceding year.
The effort led to substantial cost savings. The researchers estimated that a single 72-hour PPI infusion cost $378 per day, compared to about $101 to administer IV PPIs twice daily. “It's a conservative estimate because some could have switched to oral, but we couldn't make the claim [that] all would be eating or swallowing pills right away, so we used the IV,” Dr. Arora said. At an estimated cost savings of $277 per patient per day, the team determined that the intervention saved the pharmacy at least $121,000 over nine months.
One challenge was determining the rate of appropriate PPI infusions. “We did have several of our coauthors, [who] were from health administration, do these laborious chart reviews . . . so that's why we were able to get really good data,” Dr. Arora said. The trainee-led project also required a high degree of faculty involvement, particularly on the part of the GI fellowship program director, she added.
Administrative hurdles weren't an issue however, because the project won the university's Choosing Wisely challenge, a crowdsourcing competition to solve low-value problems. “We were able to make the change very quickly because you win analytic support and IT priority for your project,” Dr. Arora said.
How patients benefit
The reduction in PPI infusions carried additional tangible benefits for patients. “This allowed people to see that they didn't need to tether somebody to an IV pole for three days in the hospital,” Dr. Arora said. “That's another reason why I think the intervention spread like wildfire.”
The EHR's justification prompt is still active, and the reductions in inappropriate PPI infusion use have been sustained, said Dr. Arora. “With the accountable justification [in the EHR], we are able to monitor how often people are using it inappropriately,” she said. “We see that that number is still pretty low, so we have a relationship with pharmacy in case it comes back on their radar.”